[{"_1":2,"_71":-5,"_72":-5},"loaderData",{"_3":4,"_56":57},"root",{"_5":-7,"_6":7,"_54":55},"pocketbaseUrl","getGlobalInfoList",{"_8":9,"_52":53},"list",[10],{"_11":12,"_13":14,"_15":16,"_17":18,"_19":25,"_24":32,"_33":34,"_39":40,"_41":42,"_30":51},"collectionId","pbc_2834582692","collectionName","global_info","created","2026-01-08 05:31:18.135Z","expand",{"_19":20},"icon",{"_11":21,"_13":22,"_15":23,"_24":25,"_26":27,"_28":29,"_30":31},"pbc_3598190544","media","2026-01-12 06:39:10.626Z","id","udt22ypwp4u2mdw","img","varist_0zxcs4us0w_i9zvtk9qdn.webp","title","profile_icon","updated","2026-01-30 08:35:20.176Z","1ybnvwr6ilw2dvy","legion",{"_35":36,"_37":38},"lon",100.601093,"lat",13.705413,"name","Varist Siriumpunkul","social",[43,48],{"_44":45,"_46":47},"socialName","GitHub","link","https://github.com/BrokenHead",{"_44":49,"_46":50},"LinkedIn","https://www.linkedin.com/in/varist-siri-a64b3a213","2026-01-29 04:08:54.789Z","token","eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJjb2xsZWN0aW9uSWQiOiJfcGJfdXNlcnNfYXV0aF8iLCJleHAiOjE3Njk4NTU2NDUsImlkIjoiN3AxeWsxeTQyODV3c3RmIiwidHlwZSI6ImZpbGUifQ.74_f33iXDUSe8FvDTftzaDn15dOKrCY1L1H9drzDgzI","tz","Asia/Bangkok","routes/blog",{"_5":58,"_59":60},"https://pb-db.varist-siri.com","getBlogList",{"_8":61,"_52":53},[62],{"_11":63,"_13":64,"_15":65,"_24":66,"_67":68,"_28":69,"_30":70},"pbc_3877123827","blog","2025-12-29 18:52:09.639Z","v8lt41c6csgmpg5","richText","

Everything started about 17:00(UTC+7), I got a DM from the support that we have something weird going on while he was using internal tools. He found that when he clicks a link, it increases latency and even shows Cloudflare code 500 sometimes. So I try to help him, but we still have not found anything crucial to determine the accident since we released the new version of the company's API service, so the server may reset a bit to reconnect.
The problem is escalating, and after half an hour, our CTO is also facing the same thing, so we begin to panic (the funniest thing is that the newly released features include the new monitoring tool) after Sentry starts to ping our Email

(+ img)

After 30 minutes of sweating, I went through all the company infrastructure, release system, and visit down detector, then we know the root cause is the Cloudflare de-grading service, may us feel a bit reliefing that we don't need to role back and we go back home, since de-greade usually reslove after a while and that time is close to 7 alread, so then a make a short report to my boss CTO and on the way home

\r\n

(+ img)

The real duel is start at 19:30 (UTC+7) after I have a dinner and try to open the manga website as ususl I found it's extrimly slow until showing 500, my GF and her firend also can't open League, that time I know something is going, I open the company website that point our sweb stillaccess able in Thailand for a bit

\r\n

5 minutes later, I got a contact from the CTO and cannot access the web anymore. He told me that he got a call from the Europe(Denmark) branch about this: again, I visited the down detector and made a report to my company group about what is happening

\r\n

(+ img ?)

\r\n

At 21:30(UTC+7) after I write a posmidume report to my boss, and this acident disappear, it teach me a lot how the good code conduct, review, CI-CD testing and good monitor tool can make us present of mind, how the real single point of failure look like (some service that not use CloudFlare is working fine), in the end our team also discussing about we may need a second plan is Cloudflare did this again.

","Urgent call with Cloudflare outage!","2025-12-29 21:26:09.519Z","actionData","errors"]